Presenter: Kristi Jones
Authors: Kristi Jones
Faculty Advisor: Douglas Petersen
Institution: Brigham Young University
Most speech-language pathologists use static, norm-reference tests to diagnose language disorder. Often, these norm-referenced tests are biased against culturally and linguistically diverse students (Laing & Kamhi, 2002). Dynamic assessment has evidence of accurately identifying diverse students with a language disorder because it mitigates biases (e.g., prior language exposure) by measuring learning potential instead of current ability (Orellana et al., 2018). Traditionally, calculations of sensitivity and specificity have been obtained by using static assessment data as the reference standard (the standard used to initially confirm disorder); yet response to intervention (RTI) could be considered a more valid reference standard for language disorder. However, no dynamic assessment research has used RTI as a reference standard to initially confirm language disorder. The purpose of this study was to examine the sensitivity and specificity of a dynamic assessment of language (Dynamic Measure of Oral Narrative Discourse; DYMOND) when using an RTI reference standard. We administered the DYMOND to 255 students, which identified 22 students with a language disorder. Those students were matched with 27 typically developing students. These 49 students then received small group narrative-based language intervention over several weeks, during which interventionists determined how difficult it was for each child to respond to the intervention. Results indicated that the dynamic assessment had 79% sensitivity (accurately identifying disorder) and 94% specificity (accurately identifying no disorder), with RTI as the reference standard. We further confirmed language disorder by using a combination of the RTI data with poor performance on a nonword repetition task, weak narrative language, and an IEP as the reference standard. This resulted in 100% sensitivity and 95% specificity. These results support the use of dynamic assessment to identify language disorder. These results also highlight the extent to which sensitivity and specificity are dependent on the reference standard used to confirm language disorder.